Contents of this website © Copyright 2001 American Welding Society, All Rights Reserved ®

For Materials Inspection and Testing Personnel
About AWS Join AWS
Contact AWS Volunteer AWS
AWS Leadership AWS Sections
E-store   Buyer's Guide   Hotlinks  Consultant's Network  Careers & Classifieds
FABRICATION AND WELDING OF SOLID CRA AND CRA-CLAD MATERIALS The fabrication and welding of carbon steel or corrosion-resistant alloys (CRAs) in sour service equipment is explored


In this final article -- the last of a three-part series -- the fabrication and welding associated with carbon steel or corrosion-resistant alloys (CRAs) used in equipment exposed to sour service will be explored. The first article (Winter 2001 issue) covered the use of carbon steel and corrosion-resistant alloys in sour natural gas production. The second article (Spring 2001) covered the design and manufacture of equipment constructed from CRA-clad (metallurgically bonded) or CRA-lined (mechanically bonded) materials.

It may appear from this article that the fabrication and welding of solid CRA, CRA-clad or -lined material requires too much attention to detail to be economical, but actually the reverse is true. Investment in CRAs, when shown to be beneficial for the intended service, can lead to an increase in the safety and reliability of sour gas processing facilities, which is priceless.

Fig. 1 -- A CRA weld cladding of a raised face flange.

Extent of CRA Protection
Where corrosion-resistant alloy (CRA) material is used to clad or line equipment such as pressure vessels, heat exchangers and piping for corrosion protection in sour service, the entire internal surface must be covered with one or more layer(s) of CRA. This includes the gasket faces of any flanged connections and faces of tube-sheets and channels -- Figs. 1 and 2. Even the inside diameter of a hole drilled for the attachment of a set-on-type (O-Let) branch connection must be protected -- Fig. 3. Experience has shown partially CRA-clad or -lined equipment can be attacked by accelerated corrosive action at the exposed CRA/carbon steel interface, thereby undermining corrosion resistance and ultimately leading to a failure in service. Any internal attachments in this type of equipment should also be constructed from solid CRA material and attached directly to the carbon steel backer with CRA weld metal, to maintain the integrity of the CRA layer -- Fig. 4.

Fabrication Technique and Joint Design
Fabrication of solid CRA, CRA-clad or -lined equipment is covered in ASME Section VIII, Division 1 or 2 Rules for Construction of Pressure Vessels and in ASME B31.3 Process Piping. However, the specific fabrication technique used for a welded joint is determined by whether or not there is access to the CRA side.

Where access is possible, as in a CRA-clad pressure vessel with a manway, the following five-step technique is generally used:

Fig. 2 -- A heat exchanger tube and tube sheet.

1) Carbon steel is prepared with either a single- or double-V-groove joint design -- Figs. 5 and 6. Approximately 5/6 to 3/8 in. of CRA is removed or stripped back from the adjacent surfaces to avoid melting the CRA while welding the carbon steel side of the joint, which can lead to cracking. 2) Removal of carbon steel is confirmed by swabbing the area with a copper sulfate (blue vitriol)
solution. Differentiation between the two materials is indicated by copper plating out on the carbon steel, but not on the CRA. The recipe for this solution consists of 8 g copper sulfate, 3 mL of 1.84 specific gravity sulfuric acid and sufficient distilled water to make 500 mL, always adding the acid to the water, not vice-versa. Personal protection is very important during this procedure to avoid chemical burns.
3) The carbon steel side is welded first, using carbon steel welding consumables. The reverse side is then backgouged and welded to achieve complete penetration and fusion.
4) The CRA is restored in the stripped-back area, generally using at least two layers of CRA weld metal, although a single layer with the proper chemistry is also possible using a low- dilution welding process such as pulsed GMAW.
5) Finally, the area where the CRA has been restored is checked again to verify that there is no exposed carbon steel remaining.

In a piping system constructed of CRA-clad or -lined materials where access to the CRA side of the joint is not possible, the entire welded joint is completed using CRA weld metal. This is referred to as full alloy welding and is illustrated in Figs. 4, 7 and 8.

Fig. 3 -- A single-welded, single-bevel "O-Let" connection.

Where the CRA is mechanically bonded, as is often the case in pipeline fabrication and installation, a special joint design and seal weld are required at the CRA/carbon steel interface prior to joint fit-up -- Figs. 9 and 10. The joint design, seal weld penetration and seal weld heat input are particularly critical. Failure to use this procedure will generally result in cracking, porosity and fusion defects at this interface during subsequent welding of the joint.

Care of Base Materials
Care and preservation of solid CRA, CRA-clad or -lined materials is very important during fabrication and welding and should begin as materials are received. CRAs should not be stored outside without adequate protection. Plate clamps and other hard metallic devices should not be used in handling, unless the CRA surfaces are protected from marring. When stacking these materials, new, clean wood blocks should be used to separate each piece.

Chemical elements such as sulfur, phosphorous, lead, zinc and copper can contaminate CRA and cause embrittlement if not completely removed before welding and heat treating. These contaminants may be present in cutting fluids, grease, oil, waxes, household detergents and soaps, primers, marking crayons and temperature-indicating crayons. They must be removed by swabbing the affected area with a nonchlorinated solvent such as toluene, methyl-ethyl-ketone (MEK), acetone or isopropyl alcohol. In addition, the chemical content of the hydrotest water, cleaners, paints and inks must be carefully selected to avoid the contaminants mentioned earlier. Certification of these materials should be reviewed and maintained on file.

Fig. 4 -- A fillet-welded internal attachment.

The sulfur content of any fuel gas used to fire heat-treating furnaces must not exceed 0.5 g/100 ft3 or 141 ppm. Where heating oil is the fuel source, a low sulfur grade containing no more than 0.5% by weight, conforming to ASTM Grade 1, should be used. Acetylene gas should not be used as a fuel source for preheat, since the carbide used in its production is permitted to contain up to 0.5% sulfur by federal specifications.

The 300 series austenitic stainless steels, while classified as a CRA, are susceptible to attack by compounds containing chlorides. Possible sources of contamination are hydrostatic test water, dye penetrant solutions and even human perspiration. The chloride content of these fluids should be verified to be not greater than 50 ppm. To avoid contamination, workers should wear clean cotton gloves during handling. Similar precautions for CRAs containing 50% or more nickel are not necessary because they are resistant to this form of attack.

Storage and Handling of Consumables
The same storage and handling requirements that apply to carbon steel consumables are equally applicable to CRA consumables. The basic requirement for clean, dry storage prior to opening the manufacturer's packaging is vital. Tagging the age of each container using a "first in, first out" system is essential. GTAW filler metal should be marked with flagging identification on each end so that traceability is maintained even if the wire is cut in half. This will help prevent the inadvertent use of an incorrect CRA. The coating on CRA SMAW electrodes is just as hydroscopic (i.e., prone to moisture pickup) as conventional carbon steel low-hydrogen types and should be treated accordingly.

Fig. 5 -- A double-welded, single V-groove with equal thickness. Fig. 6 -- A double-welded, double V-groove with equal thickness.

Welding Processes and Qualification Codes
The most common welding processes used for joining solid CRA, CRA-clad or -lined materials are SMAW, GMAW, GTAW, FCAW, SAW or a combination of these processes. The qualification of a welding procedure specification (WPS) and welders to perform welding on CRAs is generally covered by ASME Section IX, Welding and Brazing Qualifications, as the referenced document in the construction codes, ASME Section VIII and ASME B31.3. However, ANSI/AWS B2.1, Specification for Welding Procedure and Performance Qualification, contains very similar rules. For offshore carbon steel pipeline construction, API 1104, Welding of Pipelines and Related Facilities, covers the qualification of WPSs and welders. But it does not address CRAs even though it is the referenced document in the relevant construction codes ASME B31.8, Gas Transmissions and Distribution Piping Systems, and ASME B31.4, Pipeline Transportation Systems for Liquid Hydrocarbons and Other Liquids. Yet there have been a number of solid CRA, CRA-clad and -lined pipelines constructed worldwide. This situation has generally been addressed in energy companies' project welding specifications by requiring qualification in accordance with both ASME IX and API 1104. In addition to the codes mentioned above, NACE MR0175, Standard Material Requirements Sulfide Stress Cracking-Resistant Metallic Materials for Oilfield Equipment, also contains service-related requirements for fabrication and welding.

Fig. 7 -- A single-welded, single V-groove with equal thickness in clad pipe.

Filler Metal and Flux Chemical Composition
From a chemical composition standpoint, selected CRA filler metal should overmatch the CRA base material in terms of the corrosion-resistant chemical elements present (generally considered to be Cr, Ni and Mo) in order to maintain corrosion resistance in the completed weld. This condition, commonly referred to as overalloying, is necessary to compensate for losses that occur during welding due to dilution of the weld metal by the base metal or elemental segregation.

If this approach is not followed, preferential corrosion of the weld can take place. In nickel-based alloys -- those nominally containing 50% or more nickel -- excessive silicon content in the deposited weld metal leads to hot-shortness and often will result in cracking. While the silicon content of nickel-based alloy GTAW, GMAW and SAW solid wire and SMAW electrodes are limited to a fairly low level by the filler metal specifications, the same cannot be said for FCAW tubular wire and SAW fluxes. In the case of SAW flux, a value known as the Basicity Index (BI) greater than 1.5 should be required, based on the well-established equation shown below:

Fig. 8 -- A single-welded, single-bevel, set-in branch connection.

In this equation, all components are in weight percent. Additionally, the SiO2 content of the SAW flux should not exceed 20% and the Si content of the as-deposited weld metal should not exceed 0.4%. There are a number of SAW flux manufacturers, both in the USA and abroad, whose products meet these requirements. As yet, there is no AWS filler metal specification covering nickel-based FCAW filler metals, so these should be selected, specified and qualified by treating the referenced manufacturer and brand name as an essential variable.

Filler Metal Mechanical Properties
When selecting filler metal in terms of the mechanical strength properties desired, those of the base metal should be matched or slightly exceeded. This approach will help satisfy the designer that the welded joint will have 100% efficiency (i.e., match the base metal). The allowable stress used in the design calculation formulas, found in ASME Section VIII, Division 1 or 2, and ASME B31.3 for vessels, exchangers and piping, are generally based upon a fraction (usually 14 to 13) of the specified minimum ultimate tensile strength (SMUTS) of a given base metal. But in pipelines or flowlines, where the design is covered by ASME B31.8, the allowable stress used in design calculation formulas is based upon a percentage (40 to 80%) of the specified minimum yield strength (SMYS) of a given base metal, also derated for design temperature.

Fig. 9 -- Joint design for CRA-lined pipe.

This matching, or slightly overmatching, approach is not usually a problem where carbon steel filler metal can be used, since each AWS carbon or low-alloy filler metal specification contains minimum tension test requirements -- tensile, yield and elongation -- for each classification. However, it is a problem where full alloy welding using CRA filler metal is required as it is in piping, pipelines and flowlines. The AWS CRA filler metal specifications for both stainless steel and nickel-based alloys contain only a tensile and elongation requirement and in some cases no mechanical property requirements at all, as shown in Table 1. The reason for the absence of such requirements, as stated in the specifications, is because the tensile property, bend ductility and soundness of the welds produced with these filler metals are generally determined during welding procedure qualification.

Variables in welding procedure (current, voltage, speed of travel), shielding medium (specific gas mixture or flux), manual dexterity of the welder, base metal composition and filler metal all influence the mechanical properties that can be achieved. However, interpretation of the mechanical properties for a welded joint as a whole is not possible using the transverse weld tensile test specimen generally required by most codes during welding procedure qualification (e.g., ASME Section IX and API 1104). This is true because the reduced section of that test specimen contains three regions (base metal, heat-affected zones and weld metal), all of which are simultaneously subjected to the same stress during testing that results in elongation and fracture of the region with lowest strength. For example, if weld metal strength is higher than that of the unaffected base metal, failure will occur outside the weld area and no quantitative information about the weld metal strength will be provided by the test. Actually, the intended purpose of the transverse weld tensile test is to verify that the welding procedure will produce welds that equal or exceed the design strength requirement, and only the ultimate tensile strength (UTS) and fracture location are reported. This apparent dilemma can be resolved, however, by including all-weld-metal tensile (AWMT) tests, where the yield strength (YS) as well as the other mechanical properties can be determined, as a supplementary requirement during welding procedure qualification. This test should be considered where the filler metal specification does not require the property upon which the design is based.

Fig. 10 -- A seal weld for CRA-line pipe.

Welding Technique and Parameters
CRA welding, confined in this article to austenitic stainless steel and nickel-based alloys, is similar to carbon steel; however, there are some noteworthy differences. For example, the voltage and amperage used for welding these CRAs should be somewhat lower and the travel speed higher than that used on carbon steel for the same filler metal size/base metal thickness combination. The resulting lower heat input will help prevent distortion and decrease the chance of compromising corrosion resistance of the completed joint and heat-affected zones. This is made necessary because the nominal coefficient of thermal conductivity (TC) for these CRAs is generally less than half that of carbon steel. Table 2, taken from ASME Section II, Part D, Material Properties, supports this statement.

The interior must be protected with an inert-gas backing, such as argon or helium, when welding piping or pipeline joints in solid CRA, CRA-clad or -lined materials from one side using GTAW or GMAW. The use of this so-called backpurge helps reduce internal oxidation that can lower the corrosion resistance of the finished joint. The backpurge also promotes good penetration, fusion and root bead shape. Although it is not appropriate to specify one maximum oxygen limit for the back-purged atmosphere to be used in each instance, a maximum of 0.5% (5000 ppm) is considered acceptable for all but the most sensitive applications (e.g., high purity pharmaceutical and food industries). The backpurge should precede arc initiation and be held steady for a period of time sufficient to achieve the low oxygen limit specified and should ordinarily be maintained until about two layers, or a minimum of 14 inch of weld metal, has been deposited.

Fig. 11 -- A single-welded, clad joint macroetch and hardness specimen. Fig. 12 -- A single-welded, clad joint chemical analysis specimen.

WPS Qualification Testing
ASME Section IX and API 1104 address WPS qualification in the United States and specify tensile and bend tests for groove welding qualification, chemical analysis and dye penetrant examination for weld cladding qualification. But for sour service, those rather limited tests are often supplemented by the following additional testing for the qualification of CRA-clad or -lined WPSs:

1) Charpy V-Notch toughness
2) Transverse weld hardness survey (Fig. 11)
3) Root bead chemical analysis (Fig. 12)
4) Corrosion tests

As mentioned earlier, the AWMT test may also be necessary, using the protocol illustrated in Figs. 13 and 14.

Fig. 13 -- An all-weld-metal tension test coupon, end view.

R. M. Denys-Labratorium Soete recently proposed an alternative to the AWMT test using a specimen called the side-grooved transverse tensile, which can be used to determine the YS of the weld metal in a completed weld joint. As illustrated in Fig. 15, removing one strip tensile specimen from the base metal and an additional one centered on a weld made in that base metal forms a base for carrying out this nonstandard test procedure. The side-groove is located in the middle of the weld on specimen A, so that failure will occur in that region. The test is conducted in the same manner as a standard tensile test. The area between the grooves is calculated from measurements taken before the tensile load is applied and test specimens are then loaded to failure, with instrumentation providing an indicated UTS and YS. The recorded YS from both test specimens (base metal and weld) are then used to calculate the yield strength ratio A/B to be used in the following formula to arrive at the actual yield strength of the weld metal (YSw).

YSw = YSp x A/B

In this equation, YSp is the actual yield strength shown on the mill test certificate for the base metal used. The recorded YS-A for the weld and B for the base metal are not representative of the actual yield properties, but they make it possible to obtain the ratio above and thus calculate the actual weld metal YS. The problem with this testing procedure is that one single test cannot determine that all welds will meet the yield strength calculated. The test can be made more representative, however, by using a heat of base metal having a YS at the higher end of the permitted range and performing at least two separate tests per WPS qualification.

Fig. 14 -- An all-weld-metal tension test coupon, side view.

Nondestructive Examination (NDE)
For volumetric examination of CRA production groove welds, radiography (RT) is generally the best choice. Ultrasonic examination (UT) of CRA production groove welds is generally ineffective because of the nature of CRA weld metal, which results in the following observed phenomena:

1) A high and variable return sound attenuation (loss) is caused by grain orientation in the weld metal.
2) An ultrasonic beam spread and/or skewing is also caused by this same grain directionality analogous to what occurs with stealth aircraft and conventional radar. The angular shapes of the fuselage do not reflect sufficient energy back to the sending unit to present an indication on the radar screen.
3) The inherent acoustic mismatch at the CRA/carbon steel interface of clad weld joints creates sound disruption and noise on the screen, which can mask an indication. Surface examination of CRA groove welds is generally conducted by dye-penetrant examination (PT) since CRAs are not sufficiently ferromagnetic to permit the use of magnetic particle examination (MT).

Penetrant testing, the copper sulfate test, as well as production chemical analysis, are generally used to examine CRA weld metal claddings, although the relevant construction codes do not specify the last two. The atomic absorption method, commonly referred to as wet chemical analysis, or optical emission spectroscopy may be used for quality control checks on the chemistry of production weld claddings.

Fig. 15 -- Side-grooved transverse tensile, A and B are yield strengths derived from notched test specimen.

LYNN C. LeBLANC, CWI, SCWI (, is Vice President and Welding/QA Specialist for ISI-Moody International, Amelia, La.

American Welding Society, 550 NW LeJeune Road, Miami, FL 33126
Phone (800) 443-9353, Intl. (305) 443-9353, Fax (305) 443-7559. Website Advertising
  • Educators
  • Expo & Convention
    Welding Journal
    Inspections Trends
    American Welder
    MTIF - Moderated Technical Inquiry Forum
  • BBS Grand Central
  • AWS Chat rooms